Mapping Anti-Asian Xenophobia: State-Level Variation in Implicit and Explicit Bias against Asian Americans across the United States

Abstract

Using national data from Project Implicit, the authors examine state-level variations in implicit and explicit bias against Asian Americans held by non–Asian Americans (n = 196,678) from 2018 to 2022. The authors also explore state-level sociodemographic correlates of both types of bias. The findings reveal considerable heterogeneity in implicit and explicit bias across states. Moreover, Republican and swing states had higher levels of implicit bias against Asian Americans, and states with older median ages and greater percentages of Asian populations were associated with less explicit bias. This study underscores the importance of state-level variation in and structural factors of biases against Asian Americans as contexts for examining attitudes toward Asian Americans.

Anti-Asian sentiment increased across the United States during the COVID-19 pandemic as a result of racialized political discourse about the virus and activation of harmful racial stereotypes of Asians, such as the “yellow peril” and perpetual foreigner (Li and Nicholson 2021). However, Asian American experiences of racism during the pandemic varied widely across regions. To improve our understanding of Asian Americans’ diverse experiences of racism, we examined state-level variations in implicit and explicit bias against the group as well as macro-level factors associated with anti-Asian sentiment at the state level. In particular, residential racial segregation, patterns of immigration, political attitudes, intergroup contact, and socioeconomic conditions have been shown to influence racial attitudes (Lee and Kye 2016). Whereas both explicit and implicit attitudes have been shown to predict intergroup behavior at an individual level (Kurtiş, Soylu Yalçınkaya, and Adams 2018), when aggregated across a geographic area, implicit attitudes are also strongly associated with aggregate levels of discrimination and inequity (Payne, Vuletich, and Lundberg 2017). Therefore, examining state-level estimates of both forms of bias is important for characterizing anti-Asian xenophobia and racism.

To visualize state-level implicit and explicit bias against Asian Americans, we used the data from Asian American implicit association test (IAT) collected between 2018–2022 by Project Implicit with state Federal Information Processing Standards codes (n = 196,678, after excluding 87,404 Asian respondents). The Asian IAT assessed the strength of automatic associations regarding the “Americanness” of Asian Americans compared with White, European Americans (Devos and Sadler 2019). Explicit bias was measured using a single item that asked participants to rate their preferences for Whites versus Asians on a seven-point scale (1 = “strong preference for Whites” to 7 = “strong preference for Asians”). We calculated state-level implicit and explicit bias in each state by averaging the individual-level scores, weighted for the population distribution by race/ethnicity in the state. Higher scores indicated more state-level implicit or explicit bias. The following state-level covariates were further examined using ordinary least squares regressions: state electoral patterns in 2020, median age, sex ratio, percentage of Asian populations, percentage of non-White populations, percentage of foreign-born Asian populations among those 18 years and older, percentage of families with income below the poverty level, and population density. Detailed information about the data collection and cleaning process is provided in the supplemental materials.

Figure 1 illustrates the state-level variability in implicit and explicit bias against Asian Americans. The mean implicit bias score was 0.31 (SD = 0.05), ranging from 0.12 (Hawaii) to 0.40 (Mississippi). The mean explicit bias score was 4.00 (SD = 0.12), ranging from 3.36 (Hawaii) to 4.18 (West Virginia). Although the percentage of non-White populations was negatively associated with both implicit and explicit bias, the percentage of foreign-born Asian adults was positively associated with both bias types. Additionally, the percentage of families below the poverty line and population density were not significantly associated with either type of bias. Residents of Republican states (β = 0.272, p = .051) and swing states (β = 0.201, p = .033) showed greater implicit associations with European Americans as “American” than Asian Americans. Conversely, older median age (β = −0.162, p = .019) and a higher percentage of Asian Americans in the state (β = −0.417, p < .001) were associated with lower levels of explicit bias against Asian Americans.

Figure 1.

State-level implicit Americanness bias and explicit bias against Asian Americans. The scores are weighted on the basis of the race/ethnicity of respondents in accordance with the American Community Survey 2021. For implicit Americanness bias, higher scores indicate that Asian Americans are viewed as more foreign compared with European Americans. For explicit bias, higher scores indicate a greater preference for White Americans compared with Asian Americans.

These findings reveal that Americans’ explicit and implicit attitudes toward Asian Americans vary across geographic contexts, with greater population diversity and Democratic states among the more robust predictors of more inclusive attitudes toward Asian Americans.

Supplemental Material

sj-docx-1-srd-10.1177_23780231231196517 – Supplemental material for Mapping Anti-Asian Xenophobia: State-Level Variation in Implicit and Explicit Bias against Asian Americans across the United States

Supplemental material, sj-docx-1-srd-10.1177_23780231231196517 for Mapping Anti-Asian Xenophobia: State-Level Variation in Implicit and Explicit Bias against Asian Americans across the United States by Nari Yoo, Harvey L. Nicholson, Doris F. Chang and Sumie Okazaki in Socius